On Tue, Mar 29, 2022 at 11:16:07AM -0700, Paul Scott wrote: > > On 3/29/22 10:32, Andrew M.A. Cater wrote:
[...] > > Again, I wouldn't use Balena Etcher - I would use dd if you have > > a Linux machine around - you can always reformat the USB drive afterwards. > > > FWIW Balena Etcher worked and my attempt with dd did not. One of the caveats with dd (or cat, or...) is that you might be tempted to pull the USB stick too early, while the write is still flushing data out. With big sticks and big RAMs the effect is more pronounced. I have the impression that I've had less failures since then (perhaps I'm of the impatient kind, who knows ;-) I'd recommend to use `oflag=sync': that means "use synchronous write", i.e. wait for each block to be flushed out before continuing. Besides I like `status=progress', which keeps you entertained while the operation is going on. I've experimented a bit with block sizes. 4k is on the small side, bigger blocks are a tad faster. For me there has been a (rather flat) sweet spot in the abouts of 64K. In total: dd if=/my/image/file of=/my/usb/stick bs=64k oflag=sync status=progress or something like that. Cheers -- t
signature.asc
Description: PGP signature